INDEX

Explanations

make do or accept

This neuron fires on words and phrases that express necessity or compulsion (e.g. “have to,” “need to,” “bit the bullet”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

觏

-1.45

-1.39

那你

-1.37

 enough

-1.34



-1.33

 lleg

-1.29

 cubre

-1.27

 utiliz

-1.27

 nasceu

-1.27

 acert

-1.27

POSITIVE LOGITS

or

1.73

and

1.53

no

1.43

》(

1.37

遢

1.35

перь

1.33

 graduación

1.30

 portátiles

1.28

even

1.25

щення

1.24

Activations Density 0.007%